Socially and professionally, most people rely upon video displays in one form or another for at least a portion of their work and/or recreation. With a growing demand for large screens and high definition television (HDTV), cathode ray tubes (CRTs) have largely given way to displays composed of liquid crystal devices (LCDs), light emitting diodes (LEDs), plasma display panels (PDPs), or front or rear projection systems.
A CRT operates by a scanning electron beam exciting phosphorous-based materials on the back side of a transparent screen, wherein the intensity of each pixel is commonly tied to the intensity of the electron beam. With an LED and plasma display, each pixel is an individual light-emitting device capable of generating its own light. With an LCD display, each pixel is a transient light-emitting device, individually adjusted to permit light to shine through or reflect through the pixel by altering the polarization of the transmitted or reflected light.
As LCD, plasma and LED screens do not utilize a large tube, as in a CRT, LCD, plasma and LED screens may be quite thin and, in many cases, are lighter than comparable CRT displays. As such, large and small flat screen displays can be provided to improve the portability of laptop computers, video displays in vehicles and airplanes, and information displays that are mounted or set in eye-catching locations.
A plurality of thin film devices, such as transistors, are typically incorporated into the screens of such flat screen devices as LCD, plasma and LED displays. Specifically, one or more transistors are commonly used to control the behavior of each pixel within the display. The individual nature of each pixel of an LED, plasma or LCD display introduces the possibility that each pixel may provide a different quantity of light. One pixel may be brighter or darker than another, a difference that may be quite apparent to the viewer.
As a flat screen display may incorporate hundreds of thousands of transistors, great care is generally applied in the fabrication of LED, plasma and LCD displays in an attempt to ensure that the pixels (and more specifically, the backplane transistors controlling the pixels) are as uniform and consistently alike as is possible. Frequently, especially with large displays, quality control measures discard a high percentage of displays before they are fully assembled. As such, displays are generally more expensive than they otherwise might be, as the manufacturers must recoup the costs for resources, time and precise tooling for both the acceptable displays and the unacceptable displays.
Traditionally, thin film devices have been formed through processes such as photolithography. In a photolithographic process, a substrate is provided and at least one material layer is uniformly deposited upon the substrate. A photo-resist layer, also commonly known simply as a photoresist, or even a resist, is deposited upon the material layer, typically by a spin coating machine. A mask is then placed over the photoresist and light, typically ultra-violet (UV) light, is applied through the mask to expose portions of the photoresist. During the process of exposure, the photoresist undergoes a chemical reaction. Generally, the photoresist will react in one of two ways.
With a positive photoresist, UV light changes the chemical structure of the photoresist so that it is soluble in a developer. What “shows” therefore goes, and the mask provides a copy of the patterns which are to remain—such as, for example, the trace lines of a circuit. Photolithography may also be considered a 2D process, in that each layer of material is deposited and then masked. Although 3D structure may be created by stacking layers patterned via the 2D process, there is no inherent alignment feature between the layers.
A negative photoresist behaves in the opposite manner—the UV exposure causes it to polymerize and not dissolve in the presence of a developer. As such, the mask is a photographic negative of the pattern to be left. Following the developing with either a negative or positive photoresist, blocks of photoresist remain. These blocks may be used to protect portions of the original material layer, or serve as isolators or other components.
Very commonly, these blocks serve as templates during an etching process, wherein the exposed portions of the material layer are removed, such as, for example, to establish a plurality of conductive rows.
The process may be repeated several times to provide the desired thin film devices. As such, new material layers are set down on layers that have undergone processing. Such processing may inadvertently leave surface defects and/or unintended contaminant particles in the prior layers.
With respect to transistors, there are two types—bottom-gate transistors and top gate transistors. Bottom-gate transistors incorporating amorphous silicon are generally more desirable then top gate amorphous silicon transistors. This is due in part to better device performance in terms of a higher electron field effect mobility and a lower off-state leakage current.
Although desirable, the fabrication of bottom-gate amorphous silicon transistors requires precise alignment between source/drain contacts and the gate electrode. In a typical bottom-gate transistor structure, a metal gate material is formed on a substrate. A desired gate electrode is then formed by a conventional photolithographic process.
Summarized, a dielectric layer is formed over the gate metal, and a layer of active material in which a channel will be formed is deposited over the dielectric layer. In many instances a contact layer, such as, for example, a-Si:H doped to be N+, is deposited over the dielectric layer prior to the deposition of a top metal layer.
Lithography, or a similar process, and subsequent etching processes are then employed to remove a section of the top metal layer and contact layer (if provided), lying roughly over the gate metal. This removal forms the gate and drain contact electrodes. Since photolithography and etching processes may introduce at least 1μ alignment error, there are overlaps between source/drain contacts and the gate electrode by design, to ensure the electrical continuity between the source and drain when the TFT channel is at an on state.
While leaving the overlaps alleviates the alignment problem, there are several drawbacks and therefore reasons to minimize the amount of overlap. For example, the overlap causes the channel to be longer than otherwise would be necessary, which in turn limits the reduction in size of the overall structure. The TFT source to drain current is proportional to the ratio of the channel width to the channel length. Reducing the overlaps shrinks the length of the transistor, and thus provides more room for other components that may be required for an eventual device.
Furthermore, and perhaps most importantly, parasitic capacitance is established between the source/drain electrode material and the gate material in the areas of overlap. This parasitic capacitance results in feed-through voltage. When the TFT is incorporated into a display backplane to control a display pixel, this may result in inadvertent turning on of the pixel. This uncontrolled behavior results in image flicker (inaccuracy in the Off-to-On transition of the TFT), and sticking (inaccuracy in the On-to-Off transition of the TFT) in the case of a display device. In the case of a sensor device, parasitic capacitance results in readout noise.
Further, due to variations in the substrate, resolution of the lithography, alignment of the lithographic mask and other factors, the overlap may vary from TFT to TFT in an array. Such variance thus permits a variance in feed-through voltage from TFT to TFT. More simply stated, the plurality of TFTs in the array will have a range of different performance factors.
Photolithography is a precise process applied to small substrates. In part, this small-scale application is due to the high cost of the photo masks. For the fabrication of larger devices, typically, rather than employing a larger and even more costly photo mask, a smaller mask is repeatedly used—a process that requires precise alignment.
As a photolithographic process typically involves multiple applications of materials, repeated masking and etching, issues of alignment between the thin film layers is of high importance. A photolithographic process is not well suited for formation of thin film devices on flexible substrates, where expansion, contraction or compression of the substrate may result in significant misalignment between material layers, thereby leading to inoperable thin film devices. In addition, a flexible substrate is not flat—it is difficult to hold flat during the imprinting process and thickness and surface roughness typically cannot be controlled as well as with glass or other non-flexible substrates.
The issue of flatness in photolithography can be a problem because the minimum feature size that can be produced by a given imaging system is proportional to the wavelength of the illumination divided by the numerical aperture of the imaging system. However, the depth of field of the imaging system is proportional to the wavelength of the illumination divided by the square of the numerical aperture. Therefore, as resolution is increased, the flatness of the substrate quickly becomes the critical issue.
With respect to the flat screen displays introduced above, use of flexible substrates for the internal backplane controlling the pixels is often desired. Such a flexible substrate can provide a display with flexible characteristics. A flexible substrate may also be easier to handle during fabrication and provide a more mechanically robust display for the user.
Hence, there is a need for a process to provide at least one thin film transistor that overcomes one or more of the drawbacks identified above.